Demographic data supporting FDA authorization of AI devices for Alzheimer disease and related dementias
JAMA Network
About The Study:
Transparency of evidence supporting Food and Drug Administration (FDA) authorization of artificial intelligence (AI)- and machine learning -based devices for Alzheimer disease and related dementias was limited, precluding effective evaluation of training and validation dataset representativeness. Disease status (i.e., dementia type and distribution), age, and sex were reported for fewer than half of devices, while race and ethnicity data were rarely disclosed, raising uncertainty about real-world generalizability and clinical accuracy in intended populations.
Corresponding Author: To contact the corresponding author, Ravi Gupta, MD, MSHP, email ravigupta@jhmi.edu.
To access the embargoed study: Visit our For The Media website at this link https://media.jamanetwork.com/
(doi:10.1001/jama.2025.12779)
Editor’s Note: Please see the article for additional information, including other authors, author contributions and affiliations, conflict of interest and financial disclosures, and funding and support.
# # #
Media advisory: This study is being presented at the Alzheimer’s Association International Conference.
Embed this link to provide your readers free access to the full-text article
JAMA Network
About The Study:
Transparency of evidence supporting Food and Drug Administration (FDA) authorization of artificial intelligence (AI)- and machine learning -based devices for Alzheimer disease and related dementias was limited, precluding effective evaluation of training and validation dataset representativeness. Disease status (i.e., dementia type and distribution), age, and sex were reported for fewer than half of devices, while race and ethnicity data were rarely disclosed, raising uncertainty about real-world generalizability and clinical accuracy in intended populations.
Corresponding Author: To contact the corresponding author, Ravi Gupta, MD, MSHP, email ravigupta@jhmi.edu.
To access the embargoed study: Visit our For The Media website at this link https://media.jamanetwork.com/
(doi:10.1001/jama.2025.12779)
Editor’s Note: Please see the article for additional information, including other authors, author contributions and affiliations, conflict of interest and financial disclosures, and funding and support.
# # #
Media advisory: This study is being presented at the Alzheimer’s Association International Conference.
Embed this link to provide your readers free access to the full-text article
Journal
JAMA
JAMA
AI-based breast cancer risk technology receives FDA Breakthrough Device designation
Software developed at WashU Medicine on accelerated path to approval
image:
A new AI-based technology developed by WashU Medicine researchers Graham Colditz, MD (left) and Shu (Joy) Jiang, PhD, analyzes mammograms to provide a more accurate prediction of an individual’s risk of developing breast cancer over the next five years.
view moreCredit: JOE TAYLOR
A new technology that harnesses AI to analyze mammograms and improve the accuracy of predicting a woman’s personalized five-year risk of developing breast cancer has received Breakthrough Device designation from the Food and Drug Administration (FDA). Developed by researchers at Washington University School of Medicine in St. Louis, the software has been licensed to Prognosia Inc., a WashU startup company.
The system analyzes mammograms to produce a risk score estimating the likelihood that a woman will develop breast cancer over the next five years. The technology is compatible with both types of mammogram imaging available: the four 2D views of the breast produced by full-field digital mammography and the synthetic 3D view of the breast produced by digital breast tomosynthesis. Importantly, the system produces an absolute five-year risk that makes it possible to compare a woman’s risk to an average risk based on national breast cancer incidence rates. This provides a meaningful estimate that is aligned with the U.S. national risk reduction guidelines, so that clinicians will know what steps to take next if a woman’s risk is elevated.
The FDA Breakthrough Device designation provides an expedited review process for full market approval in an effort to give patients and clinicians accelerated access to new medical devices. Products that receive the designation have already undergone rigorous testing and shown excellent promise in their potential to improve treatment or the diagnosis of debilitating or life-threatening conditions.
The software package, called Prognosia Breast, was developed by Graham A. Colditz, MD, DrPH, the Niess-Gain Professor of Surgery at WashU Medicine and associate director of prevention and control at Siteman Cancer Center, based at Barnes-Jewish Hospital and WashU Medicine; and Shu (Joy) Jiang, PhD, an associate professor of surgery in the Division of Public Health Sciences in the Department of Surgery at WashU Medicine. Colditz and Jiang co-founded Prognosia in 2024 in collaboration with WashU’s Office of Technology Management (OTM) and BioGenerator Ventures, the latter of which provided both financial support and business strategy expertise from Entrepreneur-in-Residence David Smoller, PhD.
The software is a pre-trained machine learning system that analyzes mammogram images and provides an estimate of how likely a patient is to develop breast cancer over the next five years, based solely on images and a woman’s age. According to the developers, Prognosia Breast estimates a person’s five-year risk of developing breast cancer 2.2 times more accurately than the standard method, which is based on questionnaires that consider factors such as age, race and family history. The system was trained on past mammograms from tens of thousands of individuals who underwent breast cancer screening through Siteman Cancer Center. Some of them went on to develop cancer, teaching the system what to look for in the earliest stages of tumor development. Such early signs of disease can’t be perceived even by a well-trained human eye.
“We’re excited about the potential of this technology to improve risk prediction and prevention of breast cancer broadly, no matter where a woman is getting screened,” Colditz said. “The long-term goal is to make this technology available to any woman having a screening mammogram anywhere in the world. No matter the type of imaging they receive, our data show the software’s potential to identify women at increased risk of developing breast cancer over the next five years, providing them with opportunities to take targeted steps to reduce that risk.”
The new device could have a large impact on risk prediction because the infrastructure is already in place to begin immediately using the software anywhere mammography is provided. Furthermore, many women already receive regular mammograms. According to 2023 survey data from the Centers for Disease Control and Prevention, more than 75% of women ages 50 to 74 reported having received a mammogram in the past two years.
Even with widespread screening, about 34% of breast cancer patients in the U.S. are diagnosed at later stages of the disease. According to the investigators, being able to assess risk up to five years in advance of the onset of cancer is likely to improve early detection, reducing the number of late-stage cancers diagnosed. Early detection has been shown to make treatment more effective and reduce deaths from breast cancer.
“Receiving a Breakthrough Device designation is a powerful validation of the extraordinary dedication and vision of this research team to improve breast cancer diagnosis and care,” said Doug E. Frantz, PhD, vice chancellor for innovation and commercialization at WashU. “It takes years of concerted effort to produce software that could quickly be integrated into the workflow of any mammography center, significantly enhancing the clinical value of routine mammograms no matter where they are provided. This is a prime example of the vital role of entrepreneurship and commercialization at WashU in transforming cutting-edge research into real-world technologies that improve patient care.”
The device produces a five-year risk score that is intended to complement, not replace, the analysis provided by radiologists, who will continue to review the mammograms following standard protocols. According to the American Society of Clinical Oncology and the U.S. Preventive Services Task Force, a five-year risk score of 3% or higher is considered elevated. According to guidelines from these organizations, women with elevated scores should be referred to specialists who can further advise them on their options for additional screening and prevention strategies.
About one in eight women in the U.S. will be diagnosed with breast cancer in their lifetime. Those found to be at elevated risk of this cancer have the option to receive more frequent screening — which may include other types of imaging, such as MRI — and in some cases may choose to take a type of chemotherapy called tamoxifen or endocrine therapy as preventive treatments. With such options available, identifying women at high risk is important so they have access to specialists who can help guide them in making these important choices.
The developers are planning a clinical trial at Siteman Cancer Center that will apply the risk score from Prognosia Breast in combination with the standard mammography screening protocols. Standard screening protocols include the review of mammograms and measures of breast density already provided to all patients. Individuals found to be at elevated risk will be referred to Siteman’s breast health specialists, who focus on helping individuals navigate the options they have for managing high breast cancer risk.
“Despite the sophistication of today’s breast imaging and its broad use for identifying existing tumors, today’s risk prediction for breast cancer is still questionnaire-based and not very good at estimating future risk,” Jiang said. “Our work has focused on filling that need for better methods. Moving to image-based risk prediction — which our studies have shown is much more accurate — has the potential to be revolutionary for patient care.”
The current FDA designation applies to the software’s analysis of mammogram images taken at a single time point. In the future, the researchers plan to update Prognosia Breast to analyze several years of mammograms from the same individual, which may further improve the accuracy of the prediction.
About Washington University School of Medicine
WashU Medicine is a global leader in academic medicine, including biomedical research, patient care and educational programs with 2,900 faculty. Its National Institutes of Health (NIH) research funding portfolio is the second largest among U.S. medical schools and has grown 83% since 2016. Together with institutional investment, WashU Medicine commits well over $1 billion annually to basic and clinical research innovation and training. Its faculty practice is consistently within the top five in the country, with more than 1,900 faculty physicians practicing at 130 locations. WashU Medicine physicians exclusively staff Barnes-Jewish and St. Louis Children’s hospitals — the academic hospitals of BJC HealthCare — and treat patients at BJC’s community hospitals in our region. WashU Medicine has a storied history in MD/PhD training, recently dedicated $100 million to scholarships and curriculum renewal for its medical students, and is home to top-notch training programs in every medical subspecialty as well as physical therapy, occupational therapy, and audiology and communications sciences.
Chinese Medical Journal study highlights the role of artificial intelligence in prostate cancer management
Researchers reveal cutting-edge algorithms streamline clinical decision-making, enabling early detection and treatment for patients with prostate cancer
image:
Researchers shed light on cutting-edge advances of AI in diagnosis, treatment, patient prognosis prediction, and molecular subtyping of PCa.
view moreCredit: Image Credit: National Institutes of Health (NIH) at Openverse Image Source Link: https://openverse.org/en-za/image/63d62643-cd8c-4245-ae44-2b090bb46c92?q=cancer+detection&p=17
Cancer is a significant global health issue affecting millions worldwide. Among the different types of cancer, prostate cancer (PCa)—which affects the prostate gland in the male reproductive system—is the second most prevalent type of cancer among men worldwide. Early screening and treatment are critical to reduce deaths associated with PCa but it is quite challenging due to the absence of symptoms in its early stages. Recent advancements in artificial intelligence (AI) are a boon for clinicians since it enables accurate detection of diseases even with limited signs.
Recently, a team of researchers led by Dr. Liang Cheng from Brown University Warren Alpert Medical School, United States of America, and Dr. Rui Chen from Shanghai Jiao Tong University School of Medicine, China, conducted a review to identify different AI-based models used in the diagnosis and treatment of PCa. The review was published online in the Chinese Medical Journal on July 09, 2025.
“In this review, we focus on AI-assisted personalized management and precision medicine for PCa patients from the aspects of pathology and imaging; summarize the cutting-edge advances of AI in PCa diagnosis, treatment, patient prognosis prediction, and molecular subtyping; and discuss the application of foundation model in PCa,” says Dr. Cheng.
Like any other cancer, detection of PCa involves a combination of tests and physical examinations. Notably, prostate-specific antigen (PSA) test and digital rectal exam are commonly used screening tools often coupled with different imaging techniques. While PSA tests are promising for early detection, only using PSA screening often leads to overdiagnosis and unnecessary biopsies, calling for better diagnostic interventions. Asian Prostate Cancer Artificial Intelligence is one such model that uses multimodal clinical parameters to optimize screening and reduce the number of unnecessary biopsies.
Galen Prostate is another AI-based model with convolutional neural networks (CNNs) for detection of PCa. Although used after the confirmation of cancer, this model helps identify the aggressiveness of the cancer cells by optimizing Gleason grading system, a part of biopsy screening which grades the stage of PCa based on pathological classification. Additionally, imaging tools like Fuzzy C-Means clustering algorithm for magnetic resonance imaging (MRI) analysis help in differentiating between cancerous and non-cancerous tumors while ProGNet and CNN-Based MRI Segmentation Models help in identifying and outlining abnormal tissue areas (lesions). Such interventions help save time and enable early and precise detection of PCa.
A critical step in managing PCa is therapy management. Each patient suffers from a different stage of cancer where the treatment needs to be tailored according to their needs. Localized PCa are usually treated with a combination of androgen deprivation therapy (ADT) and radiotherapy. While ADT does benefit some patients, it affects the quality of life of others. Multimodal Artificial Intelligence Prostate Prognostic Model is an AI model that helps identify patients who can benefit from short-term ADT while ruling out those unlikely to respond, assisting the clinicians in personalized treatment decisions.
The study also discusses other AI based models that improve precision in radiation therapy for example, random forest-based model for radiotherapy which automates the treatment parameters and Virtual Treatment Planner which enhances radiotherapy planning by optimizing treatment parameters. Additionally, there are various smart algorithms that help assess the outcome (prognosis) of the condition. One such model is Survival Quilt that provides optimized 10-year survival predictions for patients with localized PCa.
Biochemical recurrence (BCR) signifies the recurrence of PCa after treatment while metastasis marks the spread of cancer to different organs. Tools like Prostate Cancer Lymph Node Metastasis Detector and XGBoost have been reported for their accuracy in metastasis detection and BCR predictions, respectively. Lymph Node Metastases Diagnostic Model is an advanced model that can also identify micrometastases in lymph nodes.
The review also analyzes the switch of traditional AI tools with foundational tools. Foundational tools are the futuristic, versatile AI models which train on a large set of data and can perform multiple tasks at a time. While traditional task-specific AI tools have significantly improved patient care, the advent of foundational tools marks a groundbreaking shift in healthcare.
Overall, the study provides a complete update on how AI is transforming clinical practice. While the advancements show a significant success, researchers also highlight the challenges for implementation of such tools, emphasizing on further research strategies to reduce the large data requirements and AI bias.
“In the future, as databases become more robust, algorithms are further refined, and supportive laws and regulations are developed, AI is poised to play an even more transformative role in precision medicine for PCa,” concludes Dr. Chen.
***
Reference
DOI: 10.1097/CM9.0000000000003689
Journal
Chinese Medical Journal
Method of Research
Literature review
Subject of Research
Not applicable
Article Title
Artificial intelligence in prostate cancer
Article Publication Date
29-Jul-2025
